Three-dimensional (3D) bioprinting has emerged as a highly promising technology within the realms of tissue engineering and regenerative medicine. The assessment of printability is essential for ensuring the quality of bio-printed constructs and the functionality of the resultant tissues. Polymer materials, extensively utilized as bioink materials in extrusion-based bioprinting, have garnered significant attention from researchers due to the critical need for evaluating and optimizing their printability. Machine learning, a powerful data-driven technology, has attracted increasing attention in the evaluation and optimization of 3D bioprinting printability in recent years. This review provides an overview of the application of machine learning in the printability research of polymers for 3D bioprinting, encompassing the analysis of factors influencing printability (such as material and printing parameters), the development of predictive models, and the formulation of optimization strategies. Additionally, the review briefly explores the utilization of machine learning in predicting cell viability, evaluates the advanced nature and developmental potential of machine learning in 3D bioprinting, and examines the current challenges and future trends.

As a functional biomaterial, silk fibroin has been widely used in drug release, cell encapsulation and tissue regeneration. To meet the requirements of these applications, the properties of silk fibroin-based materials should be finely tunable. Many useful properties of biomaterials emerge from the collective interactions among ordered and disordered domains. Thus, increasing subtle control of silk hierarchical structures is required. As a characteristic of ordered silk fibroin, crystalline silk fibroin (CSF) is an important part of silk fibroin-based biomaterials, but the preparation of CSF solution, especially high concentration CSF solution, remains a challenge. Here, a solution composed of β-sheet-rich silk fibroin is reported. These CSF were obtained by the sonication of silk fibroin hydrogel, destroying the hydrogel network, and turning silk fibroin hydrogels into CSF solution. These β-sheet-rich CSF solutions were stable enough for several days or even weeks. In addition, they were typically ordered crystalline domains, which could be mixed with disordered domains and fabricated into porous scaffolds, films, hydrogels and other silk fibroin-based scaffolds with different properties.

The development of 3D printing has recently attracted significant attention on constructing complex three-dimensional physiological microenvironments. However, it is very challenging to provide a bio-ink with cell-harmless and high mold accuracy during extrusion in 3D printing. To overcome this issue, a technique improving the shear-thinning performance of semi-IPN bio-ink, which is universally applicable to all alginate/gelatin-based materials, was developed. Semi-IPN bio-ink prepared by cyclic heating–cooling treatment in this study can reduce the cell damage without sacrificing the accuracy of the scaffolds for its excellent shear-thinning performance. A more than 15% increase in post-printing Cell viability verified the feasibility of the strategy. Moreover, the bio-ink with low molecular weight and wide molecular weight distribution also promoted a uniform cell distribution and cell proliferation in clusters. Overall, this strategy revealed the effects of molecular parameters of semi-IPN bio-inks on printing performance, and the cell activity was studied and it could be widely applicable to construct the simulated extracellular matrix with various bio-inks.

The osteochondral interface—characterized by a steep gradient in both composition and mechanical properties—remains one of the most challenging anatomical sites to regenerate. Reconstructing this spatially complex heterogeneity continues to confound conventional osteochondral grafts. Although multilayer scaffolds are widely adopted, interfacial delamination frequently compromises repair outcomes. Here, we report a self-healing, physically crosslinked silk fibroin–based 3D-printing ink that incorporates gelatin and nano-hydroxyapatite for the fabrication of bilayer scaffolds with robust interfacial bonding. By tuning ultrasonication time of silk fibroin and gelatin content, the ink exhibits exceptional printability and cytocompatibility, enabling >90% post-printing cell viability. Leveraging a dual-nozzle alternating-print strategy, we generated bilayer constructs that display a stable interface and layerspecific mechanical heterogeneity. Both upper- and lower-layer inks possess good self-healing capacity, eliminating delamination and yielding a monolithic scaffold. Functional analyses revealed significant upregulation of the chondrogenic marker type II collagen in the upper layer and the osteogenic marker RUNX2 in the lower layer, achieving bidirectional lineage instruction required for osteochondral regeneration. This silk/gelatin-based, physically crosslinked, integrative bilayer scaffold offers a promising therapeutic platform for osteochondral defect repair.